The Spences Bridge Group is a mid-Cretaceous (104 Ma) volcanic succession in the southern Intermontane Belt of the Canadian Cordillera (50.5 ¿N, 121 ¿W). It comprises the Pimainus Formation (mafic to felsic lava, volcaniclastic and interbedded epiclastic rocks) and the overlying Spius Formation (andesitic lava flows). Including previous work, we have 55 sites distributed among 15 localities representing most of the >3000 m thickness. Forty-seven sites (286 oriented cores, 457 specimens), mainly andesites, yielded acceptable data. The beds are gently to moderately tilted, partly due to synvolcanic deformation. Samples taken from a sequence of flows at any one locality have, with one exception, well-grouped magnetization directions. Polarities are all normal, as excepted for rocks laid down in the Cretaceous Normal Superchron. Declinations always are clockwise of that expected of cratonic North America, indicating 60¿ rotation of the Spences Bridge Group as a whole. However, declinations differ from locality to locality, implying relative interlocality rotations about vertical axes. Hence inclination-only analysis has been used to estimate dispersion, mean inclination, and paleolatitude. Minimum dispersion for the Spius Formation (27 sites) was achieved after 80% untilting, but the changes between 80 and 100% untilting are insignificant, indicating that magnetization was acquired predominantly before tilting. By contrast, minimum dispersion for the Pimainus Formation (20 sites) was achieved after 50% untilting, indicating that magnetization was acquired after synvolcanic tilting when buried beneath the overlying Spius Formation.

Polished thin section studies show that magnetite in the Pimainus Formation has undergone extensive low temperature hydrothermal alteration, whereas magnetite in the Spius Formation shows little alteration. Throughout the range of tilt correction, from 0% to 100%, the mean inclinations of both formations were less than expected from observations obtained from mid-Cretaceous rocks of cratonic North America. The best estimate of paleolatitude (from the Spius Formation) is 50.8¿¿5.0¿ (P=0.05), which is 9.5¿¿5.7¿ less than would be expected had the rocks had been rigidly attached to North America. This corresponds to displacement from the south of 1100¿600 km. Displacement between the northern Intermontane Belt and craton probably was accommodated along major strike-slip faults (Northern Rocky Mountain Trench, Finlay, Pinchi, etc.). In the south, our results require a major dextral fault (the Intra-Quesnellia fault) to be situated during Late Cretaceous or Paleocene time within or marginal to the Omineca Belt, along which about 1000 km of dextral motion occurred. This could be a southern extension of the Pinchi Fault whose trace is now obscured by Eocene extension and tectonic denudation. The results also indicate that the largest tectonic discontinuity in the Canadian Cordillera occurs not to the east of the Intermontane Belt, as commonly assumed, but to the west, because the displacements relative to cratonic North America observed from the Intermontane Belt are only about one third of those observed from the Coast Belt. ¿ American Geophysical Union 1995